CryoSTAR: Leveraging Structural Prior and Constraints for Cryo-EM Heterogeneous Reconstruction

Yilai Li*, Yi Zhou*, Jing Yuan*, Fei Ye, Quanquan Gu#
ByteDance Research

*Equal Contribution     #Corresponding Author
{, zhouyi.naive, yuanjing.eugene, yefei.joyce,}

CryoSTAR is a neural network based framework for recovering conformational heterogenity of protein complexes. By leveraging the structural prior and constraints from a reference pdb model, cryoSTAR can output both the protein structure and density map.


Resolving conformational heterogeneity in cryo-electron microscopy (cryo-EM) datasets remains a significant challenge in structural biology. Previous methods have often been restricted to working exclusively on volumetric densities, neglecting the potential of incorporating any pre-existing structural knowledge as prior or constraints. In this paper, we present a novel methodology, cryoSTAR, that harnesses atomic model information as structural regularization to elucidate such heterogeneity. Our method uniquely outputs both coarse-grained models and density maps, showcasing the molecular conformational changes at different levels. Validated against four diverse experimental datasets, spanning large complexes, a membrane protein, and a small single-chain protein, our results consistently demonstrate an efficient and effective solution to conformational heterogeneity with minimal human bias. By integrating atomic model insights with cryo-EM data, cryoSTAR represents a meaningful step forward, paving the way for a deeper understanding of dynamic biological processes.

Results on synthetic data

pdb: 1ake4ake

cryoSTAR's structure
cryoSTAR's density

Results on experimental data



        @article {li2023cryostar,
          author = {Yilai Li and Yi Zhou and Jing Yuan and Fei Ye and Quanquan Gu},
          title = {CryoSTAR: Leveraging Structural Prior and Constraints for Cryo-EM Heterogeneous Reconstruction},
          elocation-id = {2023.10.31.564872},
          year = {2023},
          doi = {10.1101/2023.10.31.564872},
          URL = {},
          eprint = {},
          journal = {bioRxiv}